Absolute Localization and Mapping of Aerial Manipulators

Abstract

This chapter explains a multi-sensor SLAM (Simultaneous localization and Mapping), which creates an accurate 3D map for localization in aerial manipulation environments that are complex, dynamic and unstructured. This technique builds a multi-sensor map that contains measurements from different multi-modal sensors in the same geometrical framework, exploiting the synergies between different onboard sensors: 3D LIDAR; stereo camera; and range measurement beacons. The second part of this chapter explains a technique to provide real-time 6-DoF robot pose estimation using a MCL-based scheme that receives as input the measurements from all available on-board sensors of the aerial robot, and an enriched map built off-line, and uses this map for on-line 6-DoF localization.